How Sappi Europe is deploying AI across its mills and customer operations
Sappi Europe has implemented an advanced machine learning operations (MLOps) system to enhance production efficiency, improve sustainability performance, and deliver additional value to its customers.
Artificial intelligence is increasingly transforming industrial operations by converting technical capabilities into tangible operational improvements. However, according to analyses by consulting firms, many AI initiatives fail not due to limitations of the models themselves, but because of difficulties in integrating them effectively into everyday processes.
To address this challenge, Sappi Europe has developed a structured, data-driven approach across its production, planning, and environmental activities. The objective is to streamline operations, reduce energy costs, and improve day-to-day performance while simultaneously strengthening customer benefits.
MLOps, machine learning and AI
Machine learning, a key component of artificial intelligence, enables systems to analyse data patterns and develop models capable of automating tasks and improving operational performance. However, building and deploying these models typically requires coordination across multiple teams, extensive configuration, and significant computational resources.
MLOps addresses these challenges by introducing structured and repeatable processes for managing machine learning models. It enables the systematic handling of data, model development, testing, and deployment into daily operations. When underlying data or results change, the system automatically updates the models, ensuring consistent performance over time.
Sappi carried out its MLOps development project in partnership with Orange Business between October 2024 and June 2025. The initiative was designed to overcome issues such as fragmented data sources, reliance on manual processes, and limited real-time operational visibility, ultimately improving automation, process optimisation, and sustainability.
MLOps, energy use and sustainability
A primary focus of the initiative has been energy analytics, an area traditionally affected by fragmented data and manual control systems, which limit the ability to optimise energy consumption and reduce waste.
Through the deployment of its AI-driven MLOps platform, Sappi is now able to access standardised and automated data across its sites. This enables more efficient energy use, reduced waste generation, and more sustainable production processes.
At the Maastricht mill, for example, operations include both drawing electricity from and supplying electricity back to the Dutch grid. The introduction of MLOps has enhanced energy utility management by automating end-to-end data processes and improving the accuracy of interconnected operational and tactical decisions.
By integrating data and analytics into daily operations, the system generates new insights and increases flexibility in the use of energy assets, while improving the efficiency of utilities and energy-intensive processes.
MLOps, efficiency and customer value
The implementation of MLOps also delivers tangible benefits for customers. By embedding analytics into production decision-making, Sappi can improve forecasting accuracy and supply chain visibility, enabling clearer communication on delivery timelines and reducing operational risks.
Enhanced process and energy efficiency contribute to a more stable cost structure, helping to manage market volatility. At the same time, sustainability improvements extend throughout the supply chain, supporting customers seeking lower environmental footprint materials.
Additionally, collaboration with a manufacturer that integrates AI as a core operational capability can strengthen customers’ positioning in terms of innovation, resilience, and sustainability.
MLOps, scalability and long-term development
The MLOps framework has been designed with scalability in mind, allowing its application across multiple business areas, including sales forecasting and supply chain management. The system establishes clear governance rules, manages automated workflows, and continuously monitors performance to ensure model accuracy over time.
By centralising these processes, Sappi ensures consistent model updates and uniform implementation across different sites, facilitating rapid scaling. This approach shortens the time from concept to practical application on the production floor and promotes alignment between teams across mills.
Through this strategy, Sappi Europe is building a robust and future-oriented AI ecosystem, capable of supporting multiple business functions while delivering measurable benefits to both operations and customers.
